ide-diagnostics-mcp

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze live process data (thread dumps and CPU metrics) returned by the get_ide_diagnostics tool. This ingestion of untrusted data from the IDE process without explicit instructions for the agent to treat it as data rather than instructions creates a surface for indirect prompt injection.
  • Ingestion points: Output from the get_ide_diagnostics tool.
  • Boundary markers: Absent; the skill lacks delimiters or instructions to ignore embedded commands in the diagnostics data.
  • Capability inventory: The skill utilizes a diagnostic tool to read internal process state.
  • Sanitization: Absent; no validation or filtering is specified for the captured thread dumps or metrics.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 01:45 AM
Security Audit — agent-trust-hub — ide-diagnostics-mcp